The intrinsic complexity, variety of concepts and numerous ways to quantify landscape heterogeneity (LH) may hamper a better understanding of how its components relate to ecological phenomena. Our study is the first to synthesize understanding of this concept and to provide the state of the art on the subject based on a comprehensive systematic literature review of 661 articles published between 1982 and 2019. Definitions, terminologies and measurements of LH were diverse and conflicting. Most articles (534 out of 661) did not provide any definition for LH, and we found great variation among the studies that did. According to our review, only 10 studies measured the effects of different land-cover types on biotic or abiotic processes (functional LH). The remaining 651 studies measured physical attributes of the landscape without mentioning that different land-cover types may impact biotic and abiotic processes differently (structural LH). The metrics most frequently used to represent LH were the Shannon diversity index and proportion of land-cover type. Most metrics used as proxies of LH also coincided with those used to represent non-heterogeneity metrics, such as fragmentation and connectivity. We identify knowledge gaps, indicate future perspectives and propose guidelines that should be addressed when researching LH.